shiny-ble-hosting

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface for data ingestion via BLE.
  • Ingestion points: External data enters through WriteRequest.Data (SKILL.md), L2CapChannel.DataReceived (reference/api-reference.md), and L2CapFileRequest.FileName (reference/api-reference.md).
  • Boundary markers: The documentation suggests using BleServiceContext (SKILL.md) and characteristic subscription hooks to manage connection and transaction state.
  • Capability inventory: The skill provides instructions for filesystem write capabilities through the OpenL2CapFileServer and AcceptUpload APIs (SKILL.md, reference/api-reference.md).
  • Sanitization: The library implementation validates file names to prevent path traversal attacks, explicitly rejecting absolute paths and directory traversal references like '../' (reference/api-reference.md).
  • [SAFE]: The skill instructions and API reference follow standard development patterns for BLE peripheral development and include proper guidance on permission handling and secure file transfer practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 11:41 PM
Security Audit — agent-trust-hub — shiny-ble-hosting